a taxonomy for and analysis of multi-person-display ecosystems
Post on 30-May-2018
220 Views
Preview:
TRANSCRIPT
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
1/16
O R I G I N A L A R T I C L E
A taxonomy for and analysis of multi-person-display ecosystems
Lucia Terrenghi
Aaron Quigley
Alan Dix
Received: 10 October 2008/ Accepted: 15 March 2009 / Published online: 25 June 2009
Springer-Verlag London Limited 2009
Abstract Interactive displays are increasingly being dis-
tributed in a broad spectrum of everyday life environments:they have very diverse form factors and portability char-
acteristics, support a variety of interaction techniques, and
can be used by a variable number of people. The coupling
of multiple displays creates an interactive ecosystem of
displays. Such an ecosystem is suitable for particular
social contexts, which in turn generates novel settings for
communication and performance and challenges in own-
ership. This paper aims at providing a design space that can
inform the designers of such ecosystems. To this end, we
provide a taxonomy that builds on the size of the ecosystem
and on the degree of individual engagement as dimensions.
We recognize areas where physical constraints imply cer-
tain kinds of social engagement, versus other areas where
further work on interaction techniques for coupling dis-
plays can open new design spaces.
Keywords Human computer interaction Surface
interfaces Tabletop Gesture Interaction design
Display systems Social interaction
Display ecosystem scale
1 Introduction
Display devices are now available in a broad and diverse
spectrum of sizes, input/output capabilities, resolution,
power usage and portability. When considering a single
display its form factor typically defines its social affor-
dances [46]. A handheld display is comfortably visible by
one user at a time; a large wall display is visible by mul-
tiple users simultaneously and an interactive multi-touch
tabletop such as the DiamondTouch [14] affords the
simultaneous interaction of two to eight users with, for
example, novel photo sharing applications [2]. A display
devices physical constraints, such as real estate, orienta-
tion and mass, strongly affect the social context of inter-
action it supports and this is further constrained by the
users visual angle [22], territoriality [41], and capability to
reach content and manipulate the display device.
The design issues for each individual class of display
device are studied extensively in related work. New ways
to couple displays means novel ecosystems of interaction
can be realised. As such, more complex and dynamic
geometries of interaction can be created, where multiple
users and multiple displays are linked in the interaction.
Note that by ecosystem, we mean the complete system
of displays, people and the space in which they are placed.
As we elaborate Sect. 5, this may extend beyond the area
directly occupied by the screens themselves, for example,
in an airport an arrivals board a few metres across may be
visible to people over a large part of the arrivals hall.
L. Terrenghi (&)
Vodafone Group Services R&D,
Chiemgaustrasse 116, 81549 Munich, Germanye-mail: Lucia.Terrenghi@vodafone.com
A. Quigley
School of Computer Science and Informatics,
Complex and Adaptive Systems Laboratory,
University College Dublin, Dublin, Ireland
e-mail: aquigley@ieee.org
A. Dix
Computing Department, InfoLab21,
Lancaster University, South Drive, Lancaster, UK
e-mail: alan@hcibook.com
123
Pers Ubiquit Comput (2009) 13:583598
DOI 10.1007/s00779-009-0244-5
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
2/16
In this paper, we elaborate a taxonomy of multi-person-
display ecosystems so as to provide a tool for designers to
approach the design space of such dynamic geometries of
interaction, where multiple displays are coupled. Our
approach builds on physical and social dimensions, rather
than on the underlying technological solutions. As such, we
target designers of interactive spaces who might note be
familiar with specific middleware or protocols solutions, toprovide them with a reference tool based on human factors
(e.g., visual angle) and social engagement (e.g., oneone
vs. onemany situations, leading to intimacy vs. presenta-
tion types of social contexts).
We distinguish and describe the three main factors
affecting such geometries of interaction:
the size of the ecosystem defined by the coupled
displays
the nature of social interaction
the type of interaction technique that enables the
coupling of displays and transfer of interface elementsacross displays.
We first define the scope of our analysis: i.e., what we
mean with coupled displays (Sect. 2) and the range of the
spectrum that we address, i.e. a fluid middle in between
loosely and strictly coupled displays (Sect. 3). In Sect. 4 we
introduce our perspective of analysis, which builds on
physical attributes of the display ecosystem and on its
implications in terms of social interaction. Therefore, in
Sect. 5 we describe the key attributes of our multi-person-
display ecosystem taxonomy in terms of scale and the
nature of the social interaction. Section 6 relates the scale
of the ecosystem to the social interaction space, and
highlights how ecosystems in the same or similar scale can
afford different types of social interaction when different
coupling techniques are used. Thus, Sect. 7 details eco-
system binding methods and how they relate to social
context. Finally, in Sect. 8 we draw conclusions from our
study and its relevance in design.
2 Defining coupled displays
Numerous examples of systems with impressive graphics
flowing between screens can be seen in our everyday lives.
However, we need to understand what it means to have
coupled displays beyond the superficial surface features. In
order to understand this, we review two mundane scenarios
where displays are clearly not coupled.
At an airport you may use your phone to look at a web
version of the same schedule. This would be linked to
shared underlying data on the times of airplanes but is
designed for use in separate alternative interactions.
Alternatively suppose you are in the airport and use your
phone to upgrade your seat and then type the booking code
into a kiosk to receive your new ticket. This is linked in
terms of human interaction and shared underlying infor-
mation but there is at most a very weak link between the
interaction states of the phone and kiosk within the system.
We need to define coupling beyond such loose coupling
or coupling, which exists in ones mind only. So, at a
minimum coupled displays must:
share output and/or input components of the user
interface of a single interaction task;
have some system link between their interaction
statesit is not sufficient for the link to be in the
users mind and not sufficient that they are linked to the
same deep database.
Consider an alternative travel scenario at a railway sta-
tion. Imagine placing your mobile phone on a ticket vending
machine and navigating your diary to a particular event, you
look up to the kiosk screen and it offers a buy for this day
option referring to the day you can see on the mobile phone.After purchasing the ticket your phone has a save this
journeyoptionto putthe train andbooking informationback
into your diary. This is fairly low-key in terms of interactions
but the two displays do behave, for the duration of the
interaction, as if they were an integrated interactive system.
In other words, they have been coupled for a single inter-
action, and have established and displayed a system link.
3 The scope of analysis
There is a continuum of coupled displays from the fixed,
through the fluid middle and onto loosely coupled ones
relying on minimal shared data views or basic data exchange.
There are various forms of fixed multi-display arrange-
ments, in which displays are tightly connected but do not
allow any dynamic configuration or easy re-configuration.
These include (1) video-walls and gigapixel visualisation
displays, (2) rigidly fixed displays such as neighboring
departure and arrival screens and (3) multiple desktop
monitors, which are moveable but are normally kept in
fixed configurations suitable for personal use.
At the opposite end of the spectrum, loosely coupled eco-
systems do not rely on multiple displays to be available but
instead displays are appropriated if available and as required
to communicate and display content. An example of such
loose coupling, suitable for many-to-many social interaction,
is a mobile phone sending images via Bluetooth to a large
shared display, which then displays them automatically.
Neither extreme is the focus of this paper; instead, our
scope of analysis is on the dynamic centre of the continuum
of coupled displays or the fluid middle. This fluid middle
includes software support for easy personal, group and
584 Pers Ubiquit Comput (2009) 13:583598
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
3/16
opportunistic annexing [35] of displays to form larger
ecosystems. Applications in this space provide an enhanced
interaction experience beyond that which any one device
could afford.
In the fluid middle of the continuum then there needs to be:
a deeper understanding of the social activities, spaces,
conventions and affordances suitable for multi-persondisplay interaction (see Sect. 4);
an implicit or explicit way for users to express the
intention that displays be coupled and eventually
uncoupled (see Sect. 7.1);
system infrastructure to enable and support the con-
nection and coordination of the individual display
devices (see Sect. 7.2)
The system infrastructure will typically involve bespoke
applications and middleware, for example, a closely inte-
grated multi-player game that needs to be especially
installed on each users mobile phone [40] or a client
installed on each device to facilitate screen and controlsharing as in WeSpace [24]. We envisage a future where
users freely link display devices and engage in collabora-
tive interactions across them and that this will require open
architectures and protocols and not just one off middle-
ware solutions. An open architecture and protocols consist
of standards based APIs, well-understood data model and a
common interaction language.
4 Geometry, people, dynamic display ecosystems
Multi-person display ecosystems have a physical geometry
beyond the simple 2-dimensional grid of pixels of a single
display. For certain classes of application the geometry
is not important while for others the geometry offers new
classes of interaction. A suitable open architecture and
protocols could be aware of the relative orientation of both
displays and people in the ecosystem. This awareness could
translate into oriented views of data, varying screen reso-
lutions and the ability to extend a display by bringing
others into alignment. So, as displays are dynamically
coupled into new working geometries and people move and
interact in such ecosystems, how should we describe such
dynamic geometries of interaction? To address this issue
we look at the spatial relation between the scales of the
coupled displays and the peoples focal attention. This is
based on the users visual angle, which in turn depends on
the size of the individual displays, as we discuss below.
4.1 Users visual angle
Computer monitors are often arranged to subtend an angle
between 30 and 45 at the eye (e.g. laptop display
approximately 35 cm width at 70 cm distance and a 2400
monitor at 3000 distance). For larger displays similar
viewing angles are usually found (e.g., the Blinkenlights
installation [10]): as with any display, if the angle gets
close to 60 it becomes difficult to scan your eyes from side
to side or up and down without moving your head. There is
a minimum distance to allow the display to be seen as a
whole. Of course, if the visual angle gets smaller then theamount and detail of what can be displayed diminishes.
Hence, Teletext on a television displays approximately 1/5
of the number of characters across its width compared with
a computer monitor. As a television is usually positioned at
the opposite side of a room the characters end up sub-
tending a similar visual angle to those on the computer
monitor.
When considering the relation between the size of a
single display and the visual angle one can than determine
in Table 1.
We have extended Weisers classification of inch/foot/
yard displays [49] and used larger, albeit slightly archaic,imperial measurements including the perch, which is
5.5 yards (as used in [16]), and the chain, which is
22 yards. The use of these imperial measurements allows
us to deal with ratios of scale less than an order of mag-
nitude that better match the scales of social interaction we
are interested in.
Thus, the comfortable viewing area tends to create a
multi-person-display ecosystem whose scale approaches
one scale larger than the largest device itself. So a foot size
display tends to afford a yard scale ecosystem and a yard
size display fits best in rooms that are perch scale (e.g. a
medium meeting room).
One needs to notice, though, that depending on the
coupling technique (see Sect. 7) the geometry of the eco-
system can be elastic and the visual angle can vary. To
further clarify the relationship between the comfortable
viewing area and the scale of the ecosystem, in the fol-
lowing section we move from the scale of the single dis-
play to the scale of the multi-person-display ecosystem in
consideration of users focal attention (Fig. 1).
Table 1 The scale of single displays in relation to users visual angleand distance
Scale Example Display size Distance Angle
Inch Phone 3 cm 40 cm 4
Foot Tablet/laptop 35 cm 70 cm 28
Yard pub TV 1 m 3 m 19
Yard Tabletop 1 m 1 m 53
Perch Town centre 5 m 10 m 28
Chain Blinkenlights 20 m 50 m 23
Pers Ubiquit Comput (2009) 13:583598 585
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
4/16
5 Attributes of multi-person-display ecosystems
Based on the considerations above, we then describe two
attributes for the description of ecosystems of coupleddisplays in which people interact
The scale of multi-person-display ecosystems
The nature of social interaction.
5.1 Scale of the ecosystem
As anticipated, the scale of the ecosystem depends on the
size of the largest coupled device itself. So a perch scale
ecosystem tends to be based on yard sized displays (e.g.,
interactive tabletops coupled to large walled displays in an
interactive room [25]). Clearly, there will be exceptions tothis guideline. A mobile phone has a very small viewing
angle because the eye cannot focus closer than a person
accommodation distance, which varies from 7 cm to
200 cm over the course of a persons life. So people always
use it at foot distances from the viewer and not inch ones
although when in concert with other inch displays they may
be at inch distances from one another. Another example is
when people are inside the space of the display, i.e. when
you cannot see it all in one view (as in i-Land [44]): In this
case the size of display and size of the viewing space are
similar (in i-Land both at perch scale). Besides, the position
and orientation of displays can affect the scale of the
ecosystem: if the same two displays are aligned along one
surface, or are facing each-other, or have different orien-
tation (e.g. one is horizontally and the other one is verti-
cally mounted on the wall), they create different types of
ecosystems, implying different human movements for
interaction and affording different social contexts. Nacenta
et al. [32] look at this specific issue for adapting the
interface to the users perspective in multi-display
environments. The issues of visual angle and orientation in
groupware are also treated in [22, 41].
In the following we provide some examples to clarify
the size of the multi-person-display ecosystems.
5.1.1 Inch size ecosystem
An example for inch size ecosystem is two Tamagotchiconnected by Infra Red [28]. In this ecosystem users can
play games, exchange gifts and even have Tamababies if
the connection goes well. In inch size ecosystems the users
do not need to move their eyes to read the information from
one display to another. This is often determined by the fact
that either the displays are spatially arranged in an inde-
pendent manner (e.g., in mobile games such as the invisible
train [48]), or that the displays have an inch size and are
located in proximity to each other (e.g. GeneyTM [12]).
5.1.2 Foot size ecosystem
In foot size ecosystems the display surface requires the
users to move their eyes. Examples of this include the
coupling of displays in ConnecTable [47] or Tablet PCs in
Hinkleys Bumping and Stitching techniques [19, 20]. The
displays are arranged on the same focal plane so as to
create an extended interactive surface, which asks for users
to skim a larger real estate without changing the direction
of focus.
5.1.3 Yard size ecosystem
In yard size ecosystems users need to move their head to
view the coupled displays. This is the case with several
interactive rooms in which displays of different orientation
are coupled, thus requiring the change of focus. Examples
of this include UbiTable [42] and the Dynamo [23] system,
which combine personal devices as well as tabletops to a
large vertical display. UlteriorScape utilises multiple pro-
jectors and Lumisty film to enable a shared display to be
coupled with small view dependant screens [26].
5.1.4 Perch size ecosystem
Perch scale ecosystems are configured so that the users
definitely need to move their head from one display to
another one for interaction, and often their bodies as well.
Examples of this include the Stanford iRoom [25] and the
i-Land project [44], where several displays, in different
scales, are coupled and distributed within a room, and the
users act and move within such a dynamic geometry of
interaction. In this sense, the space can be considered as a
transducer of the interaction.
Fig. 1 Comfortable viewing area versus personal space versus screen
size
586 Pers Ubiquit Comput (2009) 13:583598
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
5/16
5.1.5 Chain size ecosystem
Chain size ecosystems imply that users need to move their
body to interact with and visualize the displays. Current
examples of these displays are often loosely linked spa-
tially. An example of this is the Blinkenlights installation
[10]. From their mobile phone, users could send animated
images as well as play Pong on the facade of a buildingwhose windows were illuminated (cf. Sect. 4.1) (Table 2).
5.2 Nature of social interaction
Coupled displays may be used in many social situations. At
one extreme is a face-to-face meeting with a single col-
league, for example, to exchange an image (the oneone
type of interaction). In contrast, many of the scenarios of
use for research systems assume a larger group collabo-
rating around a display table or large shared screen. We
will refer to these as the few (approximately three to nine
people). When displays are positioned or used in a publicor semi-public setting, which tends to be the case with
perch and chain scale ecosystems, there is also the possi-
bility of a larger audience or more loosely interacting
group, which we will refer to as the many (more than ten
people). In the case of a large public audience, they may
also differ in terms of their level of awareness of the dis-
plays themselves or that interaction is occurring. These
different levels of engagement are discussed in more detail
elsewhere in both artistic performance settings [15] and
relating to phone-large screen interactions [16].
Applications and situations vary as to whether there is
symmetric or open access to the displays, and in particular
the largest displays in the ecosystem, or whether one or
more participants have privileged access for some period.
We use this to distinguish five categories of interaction and
sharing.
5.2.1 Oneone
This is the simplest case where two colleagues or friends
exchange files or images, play a game or otherwise col-
laborate. An example of coupled displays supporting such a
kind of social interaction is the UbiTable [42], which
allows participants to link their laptop or tablet computers
to a shared horizontal display. The physical size of the
table (small end of yard scale) meant that it was ideally
suited to oneone interactions and all images and scenarios
used in the work are in this category.
5.2.2 Onefew
In many presentation settings a single presenter is
addressing a small group. In even simple PowerPoint
presentations, we see coupled displays as the presenter uses
a laptop or podium computer (foot scale) showing a pre-
senters view of the slides while the audience see the
current slide projected on the screen (yard or maybe perch
scale). More sophisticated coupling can be used in onefew
situations; for example, Pick-and-drop [36] could be used
to interact with an electronic whiteboard.
5.2.3 Fewfew
Multi-display systems have long been used for group work
or meeting collaborations. Colab [43] is an early example
for this: it allowed multiple users to edit and annotate a
screen image shared between individual displays and a
large projected display. This work demonstrated how social
protocols managed many of the apparent problems of
contention that arise. A more recent example in the meet-
ing/collaboration setting is WeSpace with support for
multi-user and multi-surface interaction [24]. In such set-
tings techniques such as hyperdragging allow users tomove objects fluidly from laptops and other devices to
table and wall displays [36]. Relate [18] is another tech-
nique to sense the relative location of participants laptops
within a meeting room in order to present a spatial view of
the people and their devices. IMPROMPTU [8], presents a
series of visual widgets to provide access to a user and
spatial view of a multi-device workspace.
5.2.4 One/fewmany
The most frequent current use of very large displays is at
open-air events where a small number of people control
what is shown on the public screens. More open installa-
tions have allowed members of the general public to act as
the one in control. For example the 2006 Six Nations
Championship virtual rugby ball and BBC Big Screen [6]
or the phone-building pong game in Blinkenlights [10]. In
both of these examples, anyone could participate but the
nature of the physical space or the game meant the par-
ticipation was limited at any moment (one person for the
virtual rugby and two for Blinkinlights Pong).
Various forms of digital graffiti or electronic post-its do
allow participants to use their mobile phone to leave virtual
messages or images in the environment to be viewed by
others later, for example, the Branded Meeting Places
Project allows users to attach annotations to places using
their mobile phones camera and image recognition [11].
While most of these use personal displays on a sequential
basis, others use public screens so that this shared digital
content is immediately visible, for example in the Wray
Photo Display villagers and visitors can upload photos
from their mobile phone to a public display in the village
post office [45].
Pers Ubiquit Comput (2009) 13:583598 587
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
6/16
Table 2 The scale of people-display ecosystems
Scale Example Related work Picture
Inch PDA-PDA GenyTM [12]
Foot TabletTablet Bumping and stitching [19]
Yard (yd) Wall Display-PDA Dynamo [23]
Perch (5.5yds) TabletopWall displayPDA i-Land [44]
Chain (22yds) Mobile-building wall Blinkenlights [10]
588 Pers Ubiquit Comput (2009) 13:583598
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
7/16
5.2.5 Manymany
In some systems many people can interact with the same
public screens simultaneously. For very large installations,
this often uses phone-based interaction (e.g., [10]): regrets
allowed people in the Cambridge Market Square to text
their lifes regrets to see them appear (anonymously) on a
large public screen [30]. A more strongly coupled exampleis tune_eile which is a table top application designed to be
installed in a public space and allow passers-by to share
their electronic music collections [33] (Table 3).
6 Relating the scale of the ecosystem to the social
interaction space
Different sizes of ecosystems can better accommodate
different contexts of social interaction. A foot scale display
creates a yard scale ecosystem allowing one or two view-
ers; a yard scale display creates a perch scale ecosystemallowing up to a small group of 39 viewers (few); perch
scale displays may allow 10100 viewers (many, e.g. a
classroom or an academic aula); and chain scale ecosys-
tems may allow many hundreds or thousands.
Table 4 shows examples of systems based on scale of
ecosystem and style of social interaction. This natural
relationship between scale and social interaction means
that the central diagonal is well populated but there are
few, or extreme examples in the top-right or bottom-left
corners.
6.1 Inch and foot scale
The relatively few examples of inch scale ecosystems are
all in oneone social settings, as they require heads to be
bent close together to see the displays, for example
GeneyTM [12]. At this scale even the interaction of three
people is difficult, as one person would see the displays
side on or upside down.
This sort of huddling is just possible at foot scale
allowing onefew or fewfew interactions as seen in the
coupling, coordination and negotiation in the proximity
regions around mobile devices in [27]. This scale of
interaction remains very intimate. An example of few
few interaction with a foot scale ecosystem is in
Savannah [17]. This involved children with GPS-enabled
PDAs playing a wide game in a chain-sized playing
field.
It is physically impossible, even for short periods, to get
many people heads together at an inch, foot or yard scale,
so the cells in the bottom left corner are completely blank.
However, small group interactions at this scale may of
course contribute to larger display ecologies, for example,
television presenters using technology together (yard scale)
which is then broadcast (much larger than chain scale), or
oneone encounters in wide games similar to Savannah.
6.2 Yard scale
Yard scale ecosystems are ideally suited for onefew andfewfew interactions around electronic whiteboards or
digital tabletops. In some cases, especially in formal
meeting rooms such as Colab [43], or offices with elec-
tronic walls such as i-Land [44], the scale of the ecosystem
may be in the smaller perch scale, with broadly similar
styles of interaction.
Yard scale ecosystems and displays may also be used for
oneone interactions, but the greater distance between the
participants allows some degree of privacy especially when
combined with personal devices. UbiTable [42] is a good
example of this: While the tabletop display used in Ubi-
Table would allow more than two participants, the softwaredivides the table into halves deliberately creating a social
situation where there is a personal (but viewable) and a
public space.
6.3 Perch and chain scaleone/fewmany
Even smaller perch scale ecosystems are too large for
effective oneone interactions, bringing to mind scenes in
films where an upper-class couple share breakfast at
opposite ends of a long dining table. We could imagine
exceptions to this general rules, such as forms of scientific
visualization where a couple of scientists gather over cal-culations on small displays and then see the results dis-
played on a gigapixel wall.
The perch scale and chain scale allow various forms of
onemany or fewmany lectures, demonstrations or
broadcasting. While there may be asymmetric access to the
largest display by the few, this does not preclude the many
having some level of more private interaction with their
own display devices; for example, in the early versions of
Classroom2000 (now eClass) Apple Newtons were used by
students to make personal annotations on the lecturers
electronic slides [1].
The sending of text messages to Blinkenlights [10] or
Regrets [30] is also in this space and, as previously noted,
straddles the boundary between onemany and many
many interaction styles as everyone has access in principle,
but only the output of one or a few can be seen at once.
This highlights the intrinsic problem of manymany
interaction, that even with the largest physical display
space, the available resolution is still no greater than a
single TV or laptop display, hence the individual contri-
butions of only a few can ever be visible.
Pers Ubiquit Comput (2009) 13:583598 589
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
8/16
Table 3 The nature of social interaction
Scale Example Related work Picture
Oneone Face-face meeting between 2 UbiTable
[42]
Onefew Presentation in meeting room Pick and Drop
[36]
Fewfew Around the table meeting Hyperdragging
[37]
Onemany Leaving digital post-its on public displays? Village Photo Display
[45]
590 Pers Ubiquit Comput (2009) 13:583598
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
9/16
6.4 Perch and chain scalemanymany
Manymany interactions at perch/chain scale are possible
in a number of situations:
(a) on a serialized basis (as in SMS to Blinkenlights) or
music sharing in Tune_eile [33].
(b) where individuals or groups can see and interact with
some fragment of a large virtual space, as is the case
with many PDA-based wide games. The Pirates!
game is a good example of this [9], as it turned the
physical space of the HUC2k conference into a virtual
ocean with islands, but allowed spontaneous seabattles between proximate users.
(c) where only some aggregate effect of the interactions
of the many is displayed in a shared form. For
example, the UniVote interactive public voting sys-
tem, which was implemented at the Lancaster
University, in the UK, allowed polls where users
could vote using a WAP interface on their phones and
see the results appear as a bar chart on public screens
in real-time [13].
The last of these reminds us of the need for some meansto limit potentially offensive, or merely irritating, interac-
tions on any large public screen. In the case of most text-to-
screen projects there is some form of moderation. In other
cases, such as UniVote, the interaction is limited to a
prescribed form so that the results fit within pre-orches-
trated bounds.
6.5 Perch and chain scaleone/few performance
The final part of the design space in Table 4 is the top right
corner of one/few interactions in large perch/chain scaleecosystems. The physical size of such an ecosystem means
that the results of oneone, onefew or fewfew interac-
tions would typically be visible to a public audience. One
can think of all this area as performance kind of social
setting, as installations that have this nature are typically
creating some form of artistic or entertainment perfor-
mance. The Blinkenlights Pong is exactly like this: While it
could be seen as oneone interaction in a chain scale
ecosystem, in fact the purpose is not to fulfill the personal
Table 4 The relation between the attributes of multi-person-display ecosystems
Social Interaction Scale
Inch Foot Yard Perch Chain
Oneone GenyTM Bumping and stitching (Allows privacy) UbiTable
Onefew Pick and drop
Fewfew Proximity Regions, Savannah Dynamo ?i-Land
Onemany Blinkenlights
Regrets
Manymany Tune_eile UniVote
Table 3 continued
Scale Example Related work Picture
Manymany Sharing music in public Tune_eile
[33]
Pers Ubiquit Comput (2009) 13:583598 591
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
10/16
work/leisure goals of the few, but to make an impression on
the many. One envisaged scenario is of a multiplayer shoot
em up game where each of the (few) players sees a heads-
up display on their mobile phone, but where a public screen
shows an overview of the game area, together with high-
definition images of the hot spots; thus the shared large
display would form part of the game play for the players and
be a sort of live digital sports coverage for the audience.
6.6 Different scales during social interaction
The above discussion highlights in several places the need
to think carefully about multiple interacting scales:
(a) the scale of individual displays and devices,
(b) the scale of the physical people-display configuration
that we have called ecosystem,
(c) the scale of the physical space in which this is set,
(d) the scale of the space for social interactions around
those devices.We have said that (b) is typically one size larger than the
largest display (a). In some of the examples, such as a
meeting room, the physical space (c) is the same as the
ecosystem (b). However, we have seen some examples
where the display ecosystem is set within a larger physical
space: an information kiosk in a train station, phone image
sharing in the open, catching prey in Savannah, or using the
UbiTable in a public space.
The social interaction space (d) is limited by the avail-
able physical space (c), and is typically similar in scale to
the ecosystem (b), but may also be either smaller or larger.
In the case of inch scale ecosystems the actual social space
is at least a few feet across including two heads and
shoulders. The larger scales of ecosystems may have
smaller group interactions within the space. In such cases,
if the larger display is part of the interaction then we tend
to get audience/bystanders as well as participants, so
arguably the social interaction space is still in the scale of
the ecosystem.
These observations show that the scale of the displays
and the ecology they create is intimately connected with
the forms of individual and social interaction that are
possible, or at least natural. The ways in which displays can
be coupled are very diverse though, and do not necessarily
require proximity. Put differently, the scale of the physical
space can have a different relation to the social interaction
space depending on the type of coupling interaction.
7 Ecosystem binding and social context
There currently exist a range of interaction methods for
binding two or more devices together. These methods can
be either single-person or multi-person and can often be
appropriated for multi-person-display ecosystems binding.
All of the methods described rely on a combination of
software components and hardware sensor technologies.
We first consider a range of interaction techniques and then
discuss how they may be more or less appropriate
depending on the social setting.
7.1 Interaction methods for binding
Any type of physical artifact including a display can be
instrumented with sensors. Many small devices with dis-
plays (phone, PDA, tablet and laptop) are already equipped
with short- and long-range radio technologies including
Bluetooth, 802.11, Infrared, GPRS, 3G and GPS. These are
radio systems, which in addition to their communication
roles can act as sensors. Collectively, such sensors can
provide accelerometer, location, physiological, proximity
and signal strength data. In order to orchestrate and control
the communication among the different networked sensors,different interaction techniques are possible. These are
described below in relation to the human body action used
to perform the binding and summarized in Table 5. These
classes and respective sub-classes are not fully mutually
exclusive; instead this table can be used as a reference
when considering different binding techniques for the
design of a particular ecosystem.
7.1.1 Synchronous human movement
The first class of interaction method relies on synchronous
human movement. Body movement with a device can form
an input that can be interpreted as a gesture such as shake,
bump, tap, bounce or rotation.
ConnecTable relies on the simple physical movement of
two displays towards each other at the same time to tem-
porally create a shared display area combining the two
previously personal ones [47] and is shown in a oneone
social context. SyncTap uses a simultaneous button press
method to authenticate and couple displays into an eco-
system [38]. ProxNet again uses a button push but relies on
the devices being in close proximity so the shared radio
spectrum can be surveyed and matched [39] and is shown
in a fewfew social context. Smart-Its Friends and Shake
Well Before Use rely on a single user simply holding two
devices together and shaking them to provide a method for
automatic authentication [21, 29]. While using acceler-
ometer data for secure device pairing, this may provide a
simple and familiar method for inch scale ecosystem for-
mation. The synchronous gesture of bumping two tablet
PCs together is demonstrated in [19] to temporarily form a
larger display ecosystem and is shown in a fewfew social
context. In Touch-And-Play (TAP) the physical act of the
592 Pers Ubiquit Comput (2009) 13:583598
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
11/16
person touching two devices causes the displays to form an
ecosystem [34] using the persons body as the transport
medium for the signals between the devices. Extensions of
this with physical handshakes could support fewfew
social context multi-display ecosystem formation.
7.1.2 Continuous action
The second class of interaction method relies on an action
or series of continuous actions between the displays to
form an ecosystem. Here body movement with a support
device such as a pen or sensor can form an input that can be
interpreted as a binding gesture.
Pick-and-drop relies on a special physical artifact, a pen
with a unique ID [36]. This allows users to pick up items
on one display by tapping and holding a button and then to
move it to another display and is shown in a fewfew social
context. Snarfing is a technique where a laser pointer is
used only to indicate the area of interest, and the contents
there are copied to the handheld [31]. With snarfing, the
binding is implicit. A stitch is a spatio-temporal gesture
that starts on one device display and ends on another which
is aligned with the first [20]. Here the displays act as a
larger work surface with the addition of interface elements
that are multi-display aware. Point & Shoot is a method for
forming a display ecosystem between a large situated dis-
play and a mobile phone [4, 5]. Visual codes are used to
determine both phone position with respect to the larger
display and also a bluetooth password to use to initiate a
connection.
7.1.3 Rich fixed/ad hoc infrastructure
The third class of interaction method relies on explicit user
action followed by infrastructure support binding devices
and their displays into an ecosystem.
The iRoom represents a heavily pre-configured display
eco-system that a new display can minimally couple to via
the iCrafter system in the iRos operating system [25] and is
shown in a fewfew social context. A map or controller
interface displays the geometric arrangement of screens
and lights in the room allowing application level coupling
or end user coupling to occur. Other systems, such as ARIS
[7] and WIM (World in Miniature) [50] also use map-like
or geometric views of displays within displays (in the
former iconic, in the latter as facsimile miniatures), but for
configuring applications once the displays have been cou-
pled by other means.
Table 5 Interaction methods and their sub-classes of display ecology binding with examples
Interaction Method Sub-class Examples Ecosystem scale
shown
Social context of
examples shown
Synchronous co-located human
movement (SHM)
Movement
towards
ConnecTable [47]
Sensed Proximity Regions [27]
Foot
Foot
One to one
Few to few
Button
Push
SyncTap [38]
ProxNet [39]
Upto Yard
Upto Perch
One to one
Few to fewShake Smart-Its Friends [21]
Shake Well Before Use [29]
Inch
Inch
Personal (suitable for one to one)
Personal (suitable for one to one)
Bump Tablet PC bumping [19] Foot One to few
Touch Intra-Body Comms [52] Yard One to one
(Card exchange)
Continuous action (CA) Point Tap Pick-and-drop [36] Foot/Yard One to one
PointRight iRoom [25] Perch Few to few
Gesture Snarfing [31] Chain One to few
Action & Infrastructure
(A&I)
Stroke Stitching [20] Foot One to few
Alignment Point & Shoot [4, 5] Yard One to few
Request Bluetooth Proximity Regions [27]
UbiTable [42]
Pong [10]
Foot
Yard
Chain
Few to few
One to one
Many to many
Map iRoom & iRos [25]
Relate [18]
IMPROMPTU [8]
Perch
Perch
Perch
Few to few
Few to few
Few to few
Placement BlueTable [51] Yard Few to few
Tokens TranSticks [3] Perch Many to many
Pers Ubiquit Comput (2009) 13:583598 593
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
12/16
The UbiTable allows users to couple laptop computers
with a large shared interactive surface (DiamondTouch)
[42]. Relying on a close proximity to communicate on IR
then switching to wireless, the UbiTable couples personal
record space on the laptop display with a mirrored space on
the surface. BlueTable couples small mobile devices that
have screens and wireless connections, with larger inter-
active surfaces [51]. This display ecosystem is formedusing a computer vision based handshaking method. Here
the interactive surface forms an extended interface for the
mobile device along with the system tracking its location
on the surface. TranSticks are physical tokens for virtual
connections where a user places a memory-key or memory-
stick into the appropriate reader in each device [3]. The
sticks have a shared data space that can be rendered as a
shared display between the two devices and displays and is
shown in onemany and manymany contexts.
7.2 Ecosystem binding for social contexts
Table 6 locates several of the coupling techniques described
within the same social/scale diagram as Table 4. We can
instantly see some patterns. The blank space at the lower left
is explained by the fact that there are few multi-display
ecosystems in this portion of the design space as discussedin
Sect. 4. On the whole more infrastructure techniques are used
at the larger scale and more physical/movement based
techniques (SHM & CA) at the smaller scales, reflecting the
difficulties of traversing large spaces, moving or otherwise
directly manipulating large displays and involving large
numbers of people in movement based activity.
As noted, there are many forms of multi-display systems
both wired and wireless that weld the displays into a fixed
ecology suitable for a single purpose use. Many of the
binding methods described in Sect. 7.1 are not attuned for
multi-person-display ecosystems such as Shaking or Pong
while others such as Stitch and UbiTable are. The fluid
middle of our design space requires support for connections
that are spontaneous and for ecosystem formation that lasts
as long as the real-world activity (meeting, conversation,
game, etc.) requires it.
It is clear that the infrastructure-based techniques can be
used in any social and spatial setting. They are very gen-
eric, but involve less direct physical actions. Other inter-
action techniques depend on more physical or directcoupling and so it is interesting to consider how they may
perform in the different social settings introduced in Sect.
5.2 (which loosely correlate with size of ecosystem).
7.2.1 Oneone
The exchange of business cards using IR or Bluetooth is
classic examples of one to one social interaction with
multiple loosely coupled displays. From a research per-
spective this section of the design space for both ecosystem
formation and interaction is well researched. Translating
these methods which are technologically sound into inter-action styles people would be comfortable using in day to
day one on one social settings requires further HCI
research and careful interaction design.
One issue is related to the level of intimacy required by
different techniques. Most require a level of proximity, but
not more than would be normal in oneone conversation.
However, Touch-And-Play (TAP) [34] is a more intimate
form of connection (as would be any technique using
personal area networks, or other physical contact). There
are formalized situations where we accept physical touch,
for example, a handshake, but new social conventions
would be required to allow a business contact to touch
ones phone or PDA to establish a link.
Some forms of interaction seem potentially damaging
(e.g., bumping or shaking) and again it may require a level
of mutual trust to let a stranger bump their mobile phone
against ones new tablet PC.
7.2.2 Onefew
In onefew social settings, the display ecosystem methods
for both Chain and Inch have not been explored. Clearly
the ability to form an inch scale ecosystem suitable for
interaction is limited by the physical setting.
A onefew interaction may often involve the presenter
taking up a physical location close to a large (Yard or
Perch) screen. In such cases techniques that involve some
form of synchronous human movement or continuous
action are possible between the large screen and presenters
own personal device; for example, tapping, or touching the
displays.
Where the audience or listeners (the few) also have
displays they wish to join to the ecosystem (e.g., to share
Table 6 Relationship between social interaction support and multi-
person-display ecosystem scale
Many-Many
One-Many
Few-Few
One-Few
One-One
ChainPerchYardFootInchScaleSocial
Interaction
TranSticks
Pong
Pong
Map
ProxnetStitch
Proxnet
Point-Tap
BlueTable
Point&ShootBump
ConnectableShakeUbiTable
Touch/Push
594 Pers Ubiquit Comput (2009) 13:583598
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
13/16
personal notes),thenit would also be possible for them to use
these physical means of connection on a one-by-one basis
either to the larger display or to the presenters display
device.
7.2.3 Fewfew
Some of the techniques using simultaneous human move-ment or continuous action are possible for small numbers
of participants; for example, in Savannah simultaneous
button press is used to attack prey in the game and it
would be possible to use similar techniques for coupling.
However, these become cumbersome as the number grows
beyond two or three participants and so is perhaps only
suited for games or playful interaction.
For larger groups one can envisage paradigms where
one or two people initiate an ecosystem, using any form of
oneone coupling, and then others join by coupling with
existing members of the ecosystem, rater like the situation
described for onefew interactions.
7.2.4 Onemany
In large lecture theatres or outdoor settings, this situation is
similar to that described for onefew interactions, except it
is likely that any form of large shared display is out of
reach (Perch or Chain sized display). However, in such
cases there is often the possibility of some form of control
location such as a lectern in a lecture theatre and forms of
physical coupling can be used either simply plugging in a
personal device, or by using one of the SHM or CA tech-
niques with a token at the lectern.If the audience need to connect displays then conta-
gion techniques such as those described for fewfew
interaction are possible: rather like passing a pile of notes
back through a lecture theatre, the presenter can touch,
bump or otherwise connect to a few people near the front
and then this can be passed back through the crowd.
7.2.5 Manymany
In this case, and also in theonemany situation if the audience
wish to connect their own displays into the ecosystem, simply
having large numbers of people troop to a shared display orcontrol location will usually be impractical. Infrastructure
approaches, which are possible at all scales and situations,
come into their own and many of these can be applied at this
largescale; for example,URLs or othertext tags to be typed in,
visual codes, and image recognition. However, if more
physicalor direct interaction is desired,in a gamefor example,
then contagion techniques can again be used.
In this section, we have discussed situations where there
is at most one large display and none of the coupling
techniques we have reviewed so far have been applied to
couplings of larger displays. This is partly because larger
displays are typically fixed within their environment and
likely to have hard-wired couplings. However, there are
circumstances where groups may want to connect public
display, for example, in a bar setting linking a wall display
and tabletop display. The emergence of micro-projectors
that can be embedded in mobile devices opens novelinteresting possibilities for the design of coupling tech-
niques for larger, mobile displays.
8 Conclusions and open discussion
In this paper, we have provided some criteria for critically
approaching the design of multi-person-display ecosys-
tems, in which multiple displays are coupled. Our approach
has considered physical constraints, human focus of
attention, social context and type of body movement/
interaction. This allows for the description of the designspace independently from the different technical imple-
mentation solutions for a coupling mechanism. This exer-
cise has provided some insights that we feel are relevant for
the designers of multi-person-display ecosystems:
we have shown how the scale of each single display has
an impact on the scale of the ecosystem and on the
scale of the social interaction space. In this sense, the
physical space still has an impact on the social
affordances of the ecosystem, as in the case of single
displays. However, as a result, the geometries of
interaction become more complex and dynamic; the scale of the ecosystem tends to create natural
levels of interaction, with one-to-one sharing princi-
pally at foot/yard scales, onefew and fewfew sharing
at yard or smaller perch scale and onemany and
manymany at perch/chain scale (as shown in Table 4);
to some extent, the type of coupling technique,
distinguished in terms of human body movement, can
affect the natural relationship between the scale of the
ecosystem and social interaction. In particular, accord-
ing to the type of coupling technique, interaction
geometries can become dynamic and be more or less
suitable for specific types of social contexts (e.g.gaming, performance, presentation). For example,
note that in Savannah [17] the inch scale and foot scale
ecosystems are dynamically formed and reformed for
short periods. This appears to be a pattern for a genre of
ecosystems where chain scale wide games or serious
pursuits are performed using digitally connected indi-
vidual displays interspersed with short periods of
heads-together interaction over closely coupled display
ecologies at the foot or yard scale.
Pers Ubiquit Comput (2009) 13:583598 595
123
-
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
14/16
Thus, understanding and ordering the dimensions that
affect such geometries of interaction, i.e. scale of the eco-
system, nature of social interaction and coupling interaction
technique, can help to approach the design space in a more
informed manner. Clearly, other aspects including duration
and parallelism versus serial interaction need to be consid-
ered, although these were not in the focus of this paper. We
have seen howin general synchronous human movement and
continuous action are ideally suited for oneone couplingsbut can be extended to create more complex ecosystems
using forms of contagion where a series of oneone cou-
plings gradually builds up a larger group of coupled displays.
For ecosystems involving small numbers of people and
devices, then patterns where there is a single individual or
display (e.g. a public screen) which acts as a hub to which all
other displays connect (see Fig. 2i) tends to emerge. Where
thenumberis larger,more complex patterns are neededusing
indirect connections. In onemany situations or where the
interaction is focused on a single display, this is likely to start
with a single individual or display and then spread outwards
(Fig. 2ii). Where there is no obvious leader or focus, thenecosystems may grow from several initial seeds with groups
of connected displays merging through oneone couplings
(see Fig. 2iii). We expect that such considerations in time
can inform future work for the characterization of the
taxonomy.
By looking at the coupling techniques at the pragmatic
level, i.e., how they imply peoples movement, focus and
spatial arrangementwe can assume a user-centered per-
spective to analyze and design multi-person-display eco-
system which consider the specific task and social context.
Consider for example, multiple users who are willing to
share on an interactive tabletop some of their personalcontacts stored in their mobile devices. In this context, it
makes sense to think of touch and proximity between pri-
vate and shared displays as a way to preserve a sense of
physical ownership and control of information and inter-
action. Here, the mobile device serves as a token or per-
sonal handle to private information.
By contrast, the distribution of sensors in different dis-
plays allows for interaction capabilities in a broader variety
and scale of environments, which, as represented in
Table 4, open possibilities for social interactions in perch
and chain ecosystems. In these contexts, coupling and
interaction techniques, which are based on the infrastruc-
ture and gestures in 3D, for example, can support novel
forms of computing in public spaces, which deal more with
performance and social engagement rather than personal
information management. In such scenarios, the possibili-
ties to break free from the spatial constraints of touch andproximity can afford more dynamic and elastic geometries,
which are often suitable for applications such as mobile
gaming, for example.
It is clear that as designers of such geometries we need
to tailor the different parameters, i.e., scale, social context
and coupling techniques, according to the type of experi-
ence we want to support. While doing this, we also need to
consider the culturally differences (e.g., eastern and wes-
tern) and fast changing social protocols and practice, and
how these can be supported, affected, or revolutionized by
technology.
Users of personal displays and public displays currentlyhave no expectation that their devices can seamlessly
couple as required to complete a particular task. However,
it is clear that this will change both due to the proliferation
of small personal devices with limited screen space and
also large public displays. The challenge for designers is to
develop both the systems for such coupled display eco-
systems while understanding their social context of use.
References
1. Abowd G, Atkeson C, Feinstein A, Hmelo C, Kooper R, Long S,
Sawhney N, Tani M (1996) Teaching and learning as multimedia
authoring: The Classroom 2000 Project. In the proceedings of the
ACM Multimedia96 conference, pp 187198
2. Apted T, Kay J, Quigley A (2006) Tabletop sharing of digital
photographs for the elderly. In: Proceedings of the SIGCHI
conference on human factors in computing systems, CHI 06,
ACM, pp 781790
3. Ayatsuka Y, Rekimoto J (2005) tranSticks: physically manipu-
latable virtual connections. In: Proceedings of the SIGCHI con-
ference on human factors in computing systems (Portland,
Oregon, USA, 0207 April 2005). CHI 05. ACM, New York,
pp 251260
4. Ballagas R, Rohs M, Sheridan JG (2005) Sweep and point and
shoot: phonecam-based interactions for large public displays. In:CHI 05 extended abstracts on human factors in computing sys-
tems (Portland, OR, USA, 0207 April 2005). CHI 05. ACM,
New York, pp 12001203
5. Ballagas R, Borchers J, Rohs M, Sheridan JG (2006) The smart
phone: a ubiquitous input device. IEEE Pervasive Comput
5(1):7077
6. BBC.co.uk. Rugby at the Big Screen. 16 March 2006. http://
www.bbc.co.uk/liverpool/content/articles/2006/03/16/big_screen_
rugbykick_feature.shtml. Accessed Sept 2008
7. Biehl JT, Bailey BP (2004) ARIS: an interface for application
relocation in an interactive space. In: Proceedings of graphics
Fig. 2 Contagion-based coupling building on oneone couplings: (i)
hub-and-spoke, (ii) spreading from single source, (iii) spread and
merge. In all cases the thicker lines denote earlier couplings
596 Pers Ubiquit Comput (2009) 13:583598
123
http://www.bbc.co.uk/liverpool/content/articles/2006/03/16/big_screen_rugbykick_feature.shtmlhttp://www.bbc.co.uk/liverpool/content/articles/2006/03/16/big_screen_rugbykick_feature.shtmlhttp://www.bbc.co.uk/liverpool/content/articles/2006/03/16/big_screen_rugbykick_feature.shtmlhttp://www.bbc.co.uk/liverpool/content/articles/2006/03/16/big_screen_rugbykick_feature.shtmlhttp://www.bbc.co.uk/liverpool/content/articles/2006/03/16/big_screen_rugbykick_feature.shtmlhttp://www.bbc.co.uk/liverpool/content/articles/2006/03/16/big_screen_rugbykick_feature.shtml -
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
15/16
interface 2004 (London, Ontario, Canada, May 1719, 2004).
ACM International Conference Proceeding Series, vol 62.
Canadian HumanComputer Communications Society, School of
Computer Science, University of Waterloo, Waterloo, Ontario, pp
107116
8. Biehl JT, Baker WT, Bailey BP, Tan DS, Inkpen KM, Czerwinski
M (2008) Impromptu: a new interaction framework for support-
ing collaboration in multiple display environments and its field
evaluation for co-located software development. In: Proceeding
of the twenty-sixth annual SIGCHI conference on human factors
in computing systems (Florence, Italy, 0510 April 2008).
CHI08. ACM, New York, pp 939948
9. Bjork S, Falk J, Hansson R, Ljungstrand P (2001) Pirates!using
the physical world as a game board. In: Proceedings of interact
2001, IFIP TC.13 conference on humancomputer interaction, 9
13 July 2001, Tokyo, Japan
10. Blinkenlights art installation http://www.blinkenlights.de.
Accessed Sept 2008
11. Coyne R, Williams R, Stewart J, Wright M, Lee J, Travlou P,
Boldrick S (2008) Branded meeting places: ubiquitous technol-
ogies and the design of places for meaningful human encounter.
Edinburgh University, UK. http://ace.caad.ed.ac.uk/branded/.
Accessed Sept 2008
12. Danesh A, Inkpen K, Lau F, Shu K, Booth K, Geney TM (2001)
Designing a collaborative activity for the PalmTM handheld
computer. In Proceedings of CHI01, Conference on human
factors in computing systems, Seattle, 2001, pp 388395
13. Day N, Sas C, Dix A, Toma M, Bevan C, Clare D (2007)
Breaking the campus bubble: informed, engaged, connected. In:
Procedings of BCS HCI 2007, people and computers XXI, vol 2,
BCS eWiC
14. Dietz P, Leigh D (2001) Diamond Touch: a multi-user touch
technology. In: Proceesing of UIST01, the 14th annual ACM
symposium on user interface software and technology, pp 219
226
15. Dix A, Sheridan J, Reeves S, Benford S, OMalley C (2005)
Formalising performative interaction. In: Proceedings of
DSVIS2005 (Newcastle, UK, 1315 July 2005). Lecture Notes
in Computer Science, vol 3941. Springer, Heidelberg, pp 1525
16. Dix A, Sas C (2008). Public displays and private devices: a
design space analysis. In: Workshop on designing and evaluating
mobile phone-based interaction with public displays. CHI2008,
Florence, 5 April 2008
17. Facer K, Joiner R, Stanton D, Reid J, Hull R, Kirk D (2004)
Savannah: mobile gaming and learning? J Comput Assist Learn
20:399409
18. Hazas M, Kray C, Gellersen H, Agbota H, Kortuem G, Krohn A
(2005) A relative positioning system for co-located mobile
devices. In: Proceedings of the third international conference on
mobile systems, applications, and services (Seattle, Washington,
0608 June 2005). MobiSys05. ACM, New York, pp 177190
19. Hinckley K (2003) Synchronous gestures for multiple persons
and computers. In: Proceedings of the 16th annual ACM sym-
posium on user interface software and technology (Vancouver,Canada, 0205 November 2003). UIST03. ACM, New York, pp
149158
20. Hinckley K, Ramos G, Guimbretiere F, Baudisch P, Smith M
(2004) Stitching: pen gestures that span multiple displays. In:
Proceedings of the working conference on advanced visual
interfaces (Gallipoli, Italy, 2528 May 2004). AVI04. ACM,
New York, pp 2331
21. Holmquist L, Mattern F, Schiele B, Alahuhta P, Beigl M, Gel-
lersen H (2001) Smart-its friends: a technique for users to easily
establish connections between smart artefacts, Ubicomp 2001:
Ubiquitous Computing, pp 116122
22. Inkpen KM, Hawkey K, Kellar M, Mandryk RL, Parker JK,
Reilly D, Scott SD, Whalen T (2005) Exploring display factors
that influence co-located collaboration: angle, size, number, and
user arrangement. In: Proceedings of HCI international 2005. Las
Vegas, USA, July 2005
23. Izadi S, Brignull H, Rodden T, Rogers Y, Underwood M (2003)
Dynamo: a public interactive surface supporting the cooperative
sharing and exchange of media. In: Proceedings of the 16th
annual ACM symposium on user interface software and tech-
nology (Vancouver, Canada, 0205 November 2003). UIST03.
ACM, New York, pp 159168
24. Jiang H, Wigdor D, Forlines C, Shen C (2008) System design for
the WeSpace: linking personal devices to a table-centered multi-
user, multi-surface environment. In: Proceedings of the third
annual IEEE international workshop on horizontal interactive
humancomputer systems (Tabletop) (Amsterdam, The Nether-
lands, 0103 October 2008), IEEE Computer Society, pp 105
112
25. Johanson B, Fox A, Winograd T (2002) The Interactive Work-
spaces Project: Experiences with Ubiquitous Computing Rooms.
IEEE Pervasive Computing, vol 01, no 2. AprilJune 2002, pp
6774
26. Kakehi Y, Naemura T (2008) UlteriorScape: interactive optical
superimposition on a view-dependant tabletop display. In: Pro-
ceedings of the third annual IEEE international workshop on
horizontal interactive humancomputer systems (Tabletop)
(Amsterdam, The Netherlands, 0103 October 2008), IEEE
Computer Society, pp 201204
27. Kray C, Rohs M, Hook J, Kratz S (2008) Group coordination and
negotiation through spatial proximity regions around mobile
devices on augmented Tabletops. In: Proceedings of the third
annual IEEE international workshop on horizontal interactive
humancomputer systems (Tabletop) (Amsterdam, The Nether-
lands, 0103 October 2008), IEEE Computer Society, pp 310
28. Lewis T (1997) Cars, Phones, and Tamagotchi Tribes. IEEE
Comput 11(30):142143
29. Mayrhofer R, Gellersen H (2007) Shake well before use:
authentication based on accelerometer data. In: Proceeding of the
fifth international conference on pervasive computing, Toronto,
Canada, May 2007, pp 144161
30. Mulfinger J, Budgett G, Magagnosc C (2005) Regrets [Cam-
bridge] (Accessed Oct 2008). http://www.arts.ucsb.edu/
faculty/budgett/regrets/ also see Regrets Event (Cambridge,
1/11/0520/11/05) home page (Accessed Oct 2008) http://www.
regrets.co.uk/
31. Myers BA (2001) Using handhelds and PCs together. Commun
ACM 44(11):3441
32. Nacenta MA, Sakurai S, Yamaguchi T, Miki Y, Itoh Y, Kitamura
Y, Subramanian S, Gutwin C (2007) E-conic: a perspective-
aware interface for multi-display environments. In Proceedings of
the 20th Annual ACM symposium on user interface software and
technology (Newport, Rhode Island, USA, 0710 October 2007).
UIST07. ACM, New York, pp 279288
33. OMurchu N (2008) tune_eile: A platform for social interac-tions through handheld musical devices. In: Proceedings of the
Workshop on designing multi-touch interaction techniques for
coupled public and private displays, AVI 2008, pp 5458
34. Park DG, Kim JK, Sung JB, Hwang JH, Hyung CH, Kang SW
(2006) TAP: touch-and-play. In: Grinter R, Rodden T, Aoki P,
Cutrell E, Jeffries R, Olson G (eds) Proceedings of the SIGCHI
conference on human factors in computing systems (Montreal,
Quebec, Canada, 2227 April 2006). CHI06. ACM, New York,
pp 677680
35. Pierce JS, Mahaney HE, Abowd GD (2003) Opportunistic
annexing for handheld devices: opportunities and challenges,
Pers Ubiquit Comput (2009) 13:583598 597
123
http://www.blinkenlights.de/http://ace.caad.ed.ac.uk/branded/http://www.arts.ucsb.edu/faculty/budgett/regrets/http://www.arts.ucsb.edu/faculty/budgett/regrets/http://www.regrets.co.uk/http://www.regrets.co.uk/http://www.regrets.co.uk/http://www.regrets.co.uk/http://www.arts.ucsb.edu/faculty/budgett/regrets/http://www.arts.ucsb.edu/faculty/budgett/regrets/http://ace.caad.ed.ac.uk/branded/http://www.blinkenlights.de/ -
8/14/2019 A taxonomy for and analysis of multi-person-display ecosystems
16/16
GVU Technical Report; GIT-GVU-03-31, http://hdl.handle.
net/1853/3242
36. Rekimoto J (1997) Pick-and-drop: a direct manipulation tech-
nique for multiple computer environments. In: Proceedings of the
tenth annual ACM symposium on user interface software and
technology (Banff, Alberta, Canada, October 1417, 1997).
UIST97. ACM, New York, pp 3139
37. Rekimoto J, Saitoh M (1999) Augmented surfaces: a spatially
continuous work space for hybrid computing environments. In:
Proceedings of the SIGCHI conference on human factors in
computing systems: the CHI Is the Limit (Pittsburgh, Pennsyl-
vania, USA, 1520 May 1999), CHI99. ACM, New York,
pp 378385
38. Rekimoto J (2004) Synctap: synchronous user operation for
spontaneous network connection. Pers Ubiquitous Comput
8(2):126134
39. Rekimoto J, Miyaki T, Kohno M (2004) ProxNet: secure dynamic
wireless connection by proximity sensing. In: Proceeding of the
second international conference on pervasive computing, pp 213
218
40. Sanneblad J, Holmquist LE (2003) OpenTrek: a platform for
developing interactive networked games on mobile devices. In:
Proceedings of Mobile HCI 2003, Udine, Italy
41. Scott SD, Sheelagh M, Carpendale T, Inkpen KM (2004) Terri-
toriality in collaborative tabletop workspaces. In: Proceedings of
the 2004 ACM conference on computer supported cooperative
work (Chicago, IL, USA, 0610 November 2004). CSCW04.
ACM, New York, pp 294303
42. Shen C, Everitt KM, Ryall K (2003) UbiTable: impromptu face-
to-face collaboration on horizontal interactive surfaces. In: ACM
international conference on Ubiquitous Computing (UbiComp),
October 2003, pp 281288
43. Stefik M, Bobrow DG, Foster G, Lanning S, Tatar D (1987)
WYSIWIS revised: early experiences with multi-user interfaces.
In: Proceedings of the conference on computer-supported coop-
erative work, Austin, Texas, 35 December 1986, pp 276290
(Reprinted by invitation in ACM Trans Office Inf Syst 5(2):147
167, April 1987)
44. Streitz NA, Geiler J, Holmer T, Konomi S, Muller-Tomfelde C,
Reischl W, Rexroth P, Seitz P, Steinmetz R (1999) i-Land: an
interactive landscape for creativity and innovation. In: Proceed-
ings of the CHI99 conference on human factors in computing
systems (CHI99), ACM Press, New York, pp 120127
45. Taylor N, Cheverst K, Fitton D, Race NJ, Rouncefield M, Graham
C (2007) Probing communities: study of a village photo display. In:
Proceedings of the 2007 conference of the computerhuman
interaction Special Interest Group (Chisig) of Australia on com-
puterhuman interaction: design: activities, artifacts and environ-
ments (Adelaide, Australia, 2830 November 2007). OZCHI07,
vol 251. ACM, New York, pp 1724
46. Terrenghi L (2007) Designing hybrid interactions through an
understanding of the Affordances of Physical and Digital Tech-
nologies. Published on the digital library of the Ludwig-Max-
imilians Universitat Munchen, PhD Thesis. http://edoc.ub.
uni-muenchen.de/7874/1/terrenghi_lucia.pdf
47. Tandler P, Prante T, Muller-Tomfelde C, Streitz N, Steinmetz R
(2001) Connectables: dynamic coupling of displays for the flex-
ible creation of shared workspaces. In: Proceedings of the 14th
annual ACM symposium on user interface Software and Tech-
nology (Orlando, Florida, 1114 November 2001). UIST01.
ACM, New York, pp 1120
48. Wagner D, Pintaric T, Ledermann F, Schmalstieg D (2005)
Towards massively multi-user augmented reality on handheld
devices. In: Proceedings of the third international conference on
pervasive computing (Pervasive 2005), Munich, Germany,
pp 208219
49. Weiser M (1991) The Computer for the twenty-first century. Sci
Am 265(3):94104
50. Wigdor D, Shen C, Forlines C, Balakrishnan R (2006) Table-
centric interactive spaces for real-time collaboration. In: Pro-
ceedings of the working conference on advanced visual interfaces
(Venezia, Italy, 2326 May 2006). AVI06. ACM, New York,
103107. doi:http://doi.acm.org/10.1145/1133265.1133286
51. Wilson AD, Sarin R (2007) BlueTable: connecting wireless
mobile devices on interactive surfaces using vision-based hand-
shaking. In: Proceedings of graphics interface 2007 (Montreal,
Canada, 2830 May 2007). GI07, vol 234. ACM, New York,
pp 119125
52. Zimmerman TG (1996) Personal area networks: near-field intra-
body communication. IBM Syst J 35:609617
598 Pers Ubiquit Comput (2009) 13:583598
123
http://hdl.handle.net/1853/3242http://hdl.handle.net/1853/3242http://edoc.ub.uni-muenchen.de/7874/1/terrenghi_lucia.pdfhttp://edoc.ub.uni-muenchen.de/7874/1/terrenghi_lucia.pdfhttp://doi.acm.org/10.1145/1133265.1133286http://doi.acm.org/10.1145/1133265.1133286http://edoc.ub.uni-muenchen.de/7874/1/terrenghi_lucia.pdfhttp://edoc.ub.uni-muenchen.de/7874/1/terrenghi_lucia.pdfhttp://hdl.handle.net/1853/3242http://hdl.handle.net/1853/3242
top related